A three-month field experiment by Google researchers and economist David Autor examined how AI assistance affects patent lawyers' work quality and skill development. Conducted at eleven intellectual property law firms with 133 lawyers, the study compared lawyers with access to Google's experimental AI patent-drafting tool against a control group. The results showed consistent improvements in work quality — lawyers with AI access scored 0.34 to 0.38 standard deviations higher on drafting tasks, equivalent to an 11-percentile point improvement over the control group. However, when researchers tested whether these lawyers had actually developed better judgment — using a redlining task completed without AI assistance — the results split sharply by experience level. Senior lawyers with seven or more years of experience who used AI for 90 days demonstrated measurably stronger judgment than their peers without access, suggesting the tool helped them refine professional expertise. Junior lawyers showed no average skill gain from AI use. Instead, their performance split into two groups: some scored significantly higher, while others performed notably worse. The findings raise important questions about AI's role in professional training and workforce development. While AI tools clearly boost immediate productivity and output quality, their capacity to build lasting expertise appears to depend heavily on a worker's existing knowledge base. For fields relying on learning-by-doing, the study suggests AI could create a dual-track development pattern, where the technology accelerates growth for experienced professionals while potentially leaving less-experienced workers behind without careful integration into structured mentorship programs.